Accidents between pedestrians and vehicles are, unfortunately, a fairly common occurrence. This is especially troublesome in populous, urban areas, such as large cities where the densities of motorists and pedestrians are high. As populations and population densities increase, so do the number of pedestrians, the number of motorists on the road, and the likelihood of vehicle-pedestrians accidents.
A principal factor in such vehicle-pedestrian accidents is often the failure of a motorist to detect a pedestrian. Similarly, failure of a motorist to evaluate the potential for a collision with a pedestrian increases the risk of vehicle-pedestrian accidents. Many times a pedestrian does not enter the motorist's line of sight soon enough for the motorist to avoid a collision. While there are many causes for such failures on the part of the motorist, ranging from distraction to environmental conditions, such events are undesirable regardless of their cause.
A number of pedestrian detection systems have been proposed to prevent or lessen the likelihood of vehicle-pedestrian collisions. Additionally, attempts may be made to adapt systems designed to prevent collisions generally to prevent vehicle-pedestrian collisions specifically. Many of these prior approaches are inadequate, however, as they rely on line-of-sight detection methods, or require a significant and expensive infrastructure.
Detection systems that rely on direct, line-of-sight detection methods are greatly disadvantaged in settings where numerous obstacles are present. For example, urban settings having multiple buildings, parked cars, and other visual obstacles lessen the effectiveness of such techniques by screening visible light waves used for detection. Commonly, while a first vehicle approaches a traffic intersection from a first direction, a pedestrian or another vehicle may approach the intersection from around a corner of a building or from behind a parked car, out of the direct line-of-sight of such detectors. In such a setting, these visual obstacles make it difficult for direct line-of-sight detection systems to detect pedestrians that might present a potential for collision. Therefore, detection methods that require an unobscured, line-of-sight detection path to detect a pedestrian suffer from many of the same disadvantages as the motorist.
Examples of direct line-of-sight detection systems used to prevent collisions between vehicles and pedestrians or other objects can be seen in U.S. Pat. Nos. 4,543,577 and 4,549,181 to Tachibana et al., U.S. Pat. No. 6,223,125 to Hall, U.S. Pat. Nos. 5,983,161, 6,275,773, and 6,487,500 to Lemelson et al., U.S. Patent Application Publication No. U.S. 2002/0110261 A1 to Yanai, and U.S. Patent Application Publication No. U.S. 2002/0101360 A1 to Schrage. The systems of these documents suffer the disadvantages of direct line-of-sight detection described generally above.
While some non-line-of-sight detection systems have been proposed, some of those systems rely on large infrastructures and are, therefore, only effective where the components of such infrastructures have been installed. For example, some systems are intended for use as a part of a highway sign or signal system and thus only work in places where specially outfitted signs or signals have been installed. Similarly, stationary detectors, such as cameras, inductive loop detectors, and other similar detectors, are only useful in locations where those detectors have been installed. This is disadvantageous as a large expenditure of time, effort, and money to install and maintain such an infrastructure. Also, because implementing large infrastructures universally would be difficult, they would likely only be installed in certain areas, geographically limiting the usefulness of systems relying on such infrastructures.
Examples of systems that require extensive infrastructures for detecting vehicle and/or pedestrian locations can be seen in U.S. Pat. No. 6,223,125 to Hall, U.S. Pat. No. 6,337,637 to Kubata et al., U.S. Pat. No. 6,411,328 to Franke et al., U.S. Pat. Nos. 5,983,161; 6,275,773; and 6,487,500 to Lemelson et al., U.S. Pat. No. 6,519,512 to Haas et al., and U.S. Patent Application Publication No. U.S. 2003/0016143 A1 to Ghazarian. The systems of these documents suffer the disadvantages associated with systems that make use of large infrastructures described generally above.
Accordingly, it would be desirable to develop a system and method to provide a motorist with pedestrian alerts about pedestrian locations and/or the potential for vehicle-pedestrian collisions using non-line-of-sight detection of pedestrian location, speed, and/or heading, while not requiring an extensive infrastructure. It would also be advantageous to have a system that could provide alerts to a pedestrian.